commons-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ted Dunning <ted.dunn...@gmail.com>
Subject Re: [Math]Higher precision for Significance using Math Component
Date Tue, 02 Jul 2013 17:57:01 GMT
How does significance make any sense at all at these levels?  How can
higher precision do you any good at all?  Can't you just use the underlying
z-score (or whatever) directly instead of precision?


On Tue, Jul 2, 2013 at 8:17 AM, Dahake, Nikhil <ndahake@uncc.edu> wrote:

> Hi,
>
> I am using the SimpleRegression class's getSignificance() method to
> calculate the Significance on certain data. The lowest non zero value that
> this method currently returns is : 2.220446e-16 (for the data that I am
> analyzing).
>
> However, I need a higher precision. For example, on analysis with an
> another library the lowest non zero value I got was : 9.262178e-98
>
> How can I get a very high precision with regards to significance using the
> SimpleRegression class? Please let me know.
>
> Thank you,
> Nikhil Dahake
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message